A novel soft-sensor model which incorporates PCA (principal component analysis), RBF (Radial Basis Function) networks, and MSA (Multi-scale analysis), is proposed to infer the properties of manufactured products from real process variables. PCA is carried out to select the most relevant process features and to eliminate the correlations of input variables; multi-scale analysis is introduced to acquire much more information and to reduce uncertainty in the system; and RBF networks are used to characterize the nonlinearity of the process. A prediction of the melt index (MI), or quality of polypropylene produced in an actual industrial process, is taken as a case study. Research results show that the proposed method provides promising prediction reliability and accuracy.

Authors and Affiliations

1. National Laboratory of Industrial Control Technology, Institute of Systems Engineering, Zhejiang University, Hangzhou, 310027, PRC

   Jian Shi & Xing-Gao Liu

Corresponding author

Correspondence toXing-Gao Liu.

This work is supported by the National Natural Science Foundation of China (No.20106008), the National Development and Reform Commission of China (No. Fagai Gaoji-2004-2080), and the science fund for distinguished young scholars of Zhejiang University (No.111000-581645).

Jian Shi received his BSc degree in Mechanical Engineering from Zhejiang University, China, in 2003. He is currently pursuing his PhD degree in the Institute of Systems Engineering of Zhejiang University. His research interests include olefin polymerization modeling, multivariate statistical process control, and neural networks.

Xing-Gao Liu received his PhD degree in Control Science and Engineering, from Zhejiang University in 2000. He was a Post Doctoral Fellow in the Automation Department, at Tsinghua University from 2000 to 2002. Currently, he is an Associate Professor in Process Systems Engineering, at the Department of Control Science and Engineering, Zhejiang University. His research interests include process modeling, control and optimization, fault detection and diagnosis, neural networks in olefin polymerization, and air separation and distillation processes.

Reprints and permissions